Unequal resource division occurs in the absence of group division and identity

Even though there was no explicit mention of groups or any group assignment in our individual task setup, in principle, one could imagine that, even without any instruction to do so, some participants may still project an ingroup or outgroup identity onto the single other individual when they demonstrated the same versus a different outcome. We make several comments about this possibility. First, such implicit group assignment to individuals would oppose social identity theory’s assumptions that a group manipulation is logically required to link discriminatory tendencies to intergroup processes1. Second, while this theoretical possibility cannot be excluded from our experimental design, the experimental set-up that we used does not make it likely. In Exp. 4-6, where no groups were mentioned, none out of 225 participants made any reference to groups when explaining the strategy that they used to allocate points – instead they referred mostly to disagreement or the difference in coin flip. Moreover, studies that have focussed on similarity versus difference in meaningful opinions/values/attitudes to explain discrimination8,18,21,22 have also been presented as an alternative to, rather than an extension of, the social identity literature. The a priori assumption in the discrimination domain, including by Tajfel and colleagues1, is thus that an individual difference with a single other individual cannot be taken as evidence for intergroup processes. Likewise, social congruency designs similar to those described in this manuscript have also been implemented for decades in the field of cognitive neuroscience without results being attributed to group-based or identity processes 24,27,42–46. All in all, the assumption that groups are projected to the individual remains speculative at best, given the experimental design and the assumptions of social identity theory. If intergroup processes associated with social identity theory should be expected to apply during interactions with one individual in contexts devoid of a group manipulation, then this assumption needs to be made explicit in the theory and supported by empirical evidence.

尽管在我们的个体任务设置中并未明确提及群体或任何群体分配，从原则上讲，即便没有相应的指令，一些参与者在演示相同或不同结果时，仍可能将内群体或外群体身份投射至单独的其他个体。我们对这一可能性进行了多方面的评论。首先，这种对个体的隐含群体分配与社交认同理论的假设相悖，该理论认为，逻辑上需要群体操纵才能将歧视倾向与群体间过程联系起来（1）。其次，虽然这一理论上的可能性不能被排除在我们的实验设计之外，但我们所使用的实验设置并不使其成为可能。在实验4-6中，由于未提及任何群体，225名参与者中无人提及群体，他们在解释分配分数时所使用的策略主要涉及意见分歧或硬币翻转的差异。此外，专注于相似性与差异性在有意义意见/价值观/态度中以解释歧视的研究8,18,21,22也被呈现为是对社交认同文献的替代，而非扩展。因此，在歧视领域，包括Tajfel及其同事（1）在内的先验假设是，与单个其他个体相关的个体差异不能作为群体间过程的证据。同样，与本文所述相似的社会一致性设计在认知神经科学领域已经实施了几十年，结果并未归因于基于群体的或身份的过程24,27,42-46。总之，考虑到实验设计和社交认同理论的假设，将群体投射至个体的假设至多只能是推测性的。如果社交认同理论相关的群体间过程应适用于缺乏群体操纵的个体互动中，那么这一假设需要在理论中明确提出，并由经验证据支持。